[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: What does SOAP really add?

• To: Joshua Allen <joshuaa@m...>
• Subject: Re: What does SOAP really add?
• From: Adam Turoff <ziggy@p...>
• Date: Tue, 23 Apr 2002 14:24:52 -0400
• Cc: xml-dev@l...
• In-reply-to: <4F4182C71C1FDD4BA0937A7EB7B8B4C104F05729@r...>
• References: <4F4182C71C1FDD4BA0937A7EB7B8B4C104F05729@r...>
• User-agent: Mutt/1.3.28i

On Tue, Apr 23, 2002 at 11:09:51AM -0700, Joshua Allen wrote:
> Right.  Although I think that Paul's main point was that GET is
> the easiest method to implement, and there are lots of clients that
> can do GET but not POST.  

No, Paul's main point is about idempotency.  Slacking off on a
client-side POST implementation has nothing to do with idempotency.

The intent behind GET is that it is the preferred operation to use
with idempotent resources anywhere on the web.  XSLT is a special
case, where URLs are only dereferenced through GET operations.

The issue is not that wget, XSLT, RDF, etc. are "broken", "naive"
or incapable of performing POST operations, but that some simple
SOAP requests are idempotent but cannot be addressed with GET, thus
hide an entire class of web resources.

> Paul was attempting to make the argument
> that any system which uses a POST where a GET would suffice is
> shutting out these naive clients.
>
> Whether it is the fault of the client for being naive, the fault
> of the service for requiring sophistication, or the fault of neither
> due to such usages being out of scope of the design of the service
> is a matter for endless debate.

Naivte has nothing to do with it.  

Please stop misrepresenting the issue.

Z.

• References:
  • RE: What does SOAP really add?
    • From: "Joshua Allen" <joshuaa@m...>

• Prev by Date: Re: What does SOAP really add?
• Next by Date: RE: What does SOAP really add?
• Previous by thread: Re: What does SOAP really add?
• Next by thread: RE: What does SOAP really add?
• Index(es):
  • Date
  • Thread